TargetOS Kernel Benchmarks

Semaphore Test MPC5200 ARM9 MCF5474 MCF5282 NIOS2
semPend() with waiting token 0.2 usec 1.2 usec 0.5 usec 4.0 usec 2.9 usec
semPost() with no task waiting 0.1 usec 0.9 usec 0.5 usec 5.0 usec 4.4 usec
semPost() with task made ready, no switch 1.1 usec 4.6 usec 2.5 usec 17.o usec 12.3 usec
semPost() with context switch 0.8 usec 3.7 usec 2.5 usec 21.0 usec 9.2 usec
semPend() with context switch 0.8 usec 4.1 usec 2.6 usec 20.0 usec 15.2 usec


Queue Test MPC5200 ARM9 MCF5474 MCF5282 NIOS2
quePend() with waiting message 1.0 usec 3.1 usec 1.2 usec 10.0 usec 9.7 usec
quePost() with no task waiting 0.8 usec 3.3 usec 1.7 usec 13.0 usec 15.3 usec
quePost() with task made ready, no switch 1.0 usec 3.6 usec 2.0 usec 19.o usec 12.9 usec
quePost() with context switch 0.5 usec 4.5 usec 3.7 usec 22.0 usec 13.8 usec
quePend() with context switch 0.8 usec 4.5 usec 2.6 usec 21.0 usec 20.2 usec


Nexus Test MPC5200 ARM9 MCF5474 MCF5282 NIOS2
nexPost() with no task waiting 0.5 usec 1.8 usec 1.0 usec 5.0 usec 3.8 usec
nexPost() with 1 task readied, no switch 2.6 usec 7.4 usec 3.1 usec 22.0 usec 17.6 usec
nexPost() with 2 tasks readied, no switch 2.6 usec 7.0 usec 3.7 usec 33.0 usec 13.6 usec
nexPost() from ISR with 1 > task waiting 1.3 usec 4.5 usec 3.7 usec 22.0 usec 13.8 usec
nexPost() with context switch 2.9 usec 8.5 usec 4.5 usec 35.0 usec 17.9 usec
nexPend() with context switch 1.0 usec 4.3 usec 2.5 usec 20.0 usec 14.7 usec



The results are for the following configurations:

Board: phyCORE-MPC5200B-I/O - MPC200B at 396 MHz
Compiler: GCC v4.3.2
Cache: Instruction cache enabled, data cache enabled in copy-back mode

Board: JRM X920T - ARM9 at 60 MHz
Compiler: GCC v4.3.2
Cache: Instruction cache enabled, data cache enabled in copy-back mode

Board: M5474LITE - MCF5474 at 132 MHz
Compiler: GCC v4.3.2
Cache: Instruction cache enabled, data cache enabled in write-through mode

Board: M5282LITE - MCF5282 at 64 MHz
Compiler: GCC v4.3.2
Cache: Instruction cache enabled, buffered writes enabled

Board: NEEK NIOS2/fast at 100 MHz
Compiler: GCC v4.1.2
Cache: 4KB instruction cache, 4KB copy-back data cache

For more information on kernel benchmark, please contact us:
Blunk Microsystems
3626 Fair Oaks Blvd. Suite 100
Sacramento, CA 95864
Tel: (408)323-1758
Email: sales@blunkmicro.com.